Information Analysis System, Information Analysis Method, and Information Analysis Program

1. An information analysis apparatus comprising: a time-series pattern identification unit that identifies a time-series pattern on time-series data from the time-series data including sets of at least an affirmation level on a certain proposition and a time stamp corresponding to the affirmation level, based on at least one of indicators of a temporal variation in a total number of data of the time-series data for each time interval, a temporal variation of a mean of an affirmation level for each time interval, and a temporal variation of a variance of an affirmation level for each time interval; and an effective period identification unit that identifies an effective period of the time-series data on the proposition, based on the identified time-series pattern, wherein the proposition is a statement that expresses an opinion, the affirmation level is a level of opinions stating that the certain proposition is true, and each of the time-series pattern identification unit and the effective period identification unit is a hardware device.

2. The information analysis apparatus according to claim 1 , wherein the time-series pattern identification unit further checks whether or not the time series data has assumed a burst state in which the total number of the data of the time-series data for each time interval increases in a short time to identify the time-series pattern on the time-series data.

3. The information analysis apparatus according to claim 1 , wherein the time-series pattern identification unit further checks whether or not the time-series data has assumed an inverted state in which a mean value of the affirmation levels of the time-series data for each time interval changes from a value on a negation side to a value on an affirmation side or from the value on the affirmation side to the value on the negation side to identify the time-series pattern on the time-series data.

4. The information analysis apparatus according to claim 1 , wherein the time-series pattern identification unit further checks whether or not the time-series data has assumed a divergence state in which a variance value of the affirmation level of the time-series data for each time interval exceeds a first threshold value defined in advance or the time-series data has assumed a convergence state in which the variance value falls below a second threshold value defined in advance to identify the time-series pattern on the time-series data.

5. The information analysis apparatus according to claim 1 , further comprising an effective period confidence measure calculation unit that extracts the time-series data included in the effective period identified by the effective period identification unit, and then calculates a confidence measure of the time-series data included in the extracted effective period, wherein the effective period confidence measure calculation unit is a hardware device.

6. An information analysis method comprising: identifying, by a computer, a time-series pattern on time-series data from the time-series data including sets of at least an affirmation level on a certain proposition and a time stamp corresponding to the affirmation level, based on at least one of indicators of a temporal variation in a total number of data of the time-series data for each time interval, a temporal variation of a mean of an affirmation level for each time interval, and a temporal variation of a variance of an affirmation level for each time interval; and identifying an effective period of the time-series data on the proposition, based on the identified time-series pattern, wherein the proposition is a statement that expresses an opinion, and the affirmation level is a level of opinions stating that the certain proposition is true.

7. The information analysis method according to claim 6 , comprising further checking whether or not the time-series data has assumed a burst state in which the total number of the data of the time-series data for each time interval increases in a short time to identify the time-series pattern on the time-series data.

8. The information analysis method according to claim 6 , comprising further checking whether or not the time-series data has assumed an inverted state in which a mean value of the affirmation levels of the time-series data for each time interval changes from a value on a negation side to a value on an affirmation side or from the value on the affirmation side to the value on the negation side to identify the time-series pattern on the time-series data.

9. The information analysis method according to claim 6 , comprising further checking whether or not the time-series data has assumed a divergence state in which a variance value of the affirmation level of the time-series data for each time interval exceeds a first threshold value defined in advance or the time-series data has assumed a convergence state in which the variance value falls below a second threshold value defined in advance to identify the time-series pattern on the time-series data.

10. The information analysis method according to claim 6 , comprising: extracting the time-series data included in the effective period, and then calculating a confidence measure of the time-series data included in the extracted effective period.

11. A non-transitory computer-readable recording medium storing an information analysis program that causes a computer to execute: a time-series pattern identification process that identifies a time-series pattern on time-series data from the time-series data comprising sets of at least an affirmation level on a certain proposition and a time stamp corresponding to the affirmation level, based on at least one of indicators of a temporal variation in a total number of data of the time-series data for each time interval, a temporal variation of a mean of an affirmation level for each time interval and a temporal variation of a variance of an affirmation level for each time interval; and an effective period identification process that identifies an effective period of the time-series data on the proposition, based on the identified time-series pattern, wherein the proposition is a statement that expresses an opinion, and the affirmation level is a level of opinions stating that the certain proposition is true.

12. The non-transitory computer-readable recording medium storing the information analysis program according to claim 11 , which causes the computer to execute a process that further checks whether or not the time-series data has assumed a burst state in which the total number of the time-series data for each time interval increases in a short time to identify the time-series pattern on the time-series data, in the time-series pattern identification process.

13. The non-transitory computer-readable recording medium storing the information analysis program according to claim 11 , which causes the computer to execute a process that further checks whether or not the time-series data has assumed an inverted state in which a mean value of the affirmation levels of the time-series data for each time interval changes from a value on a negation side to a value on an affirmation side or from the value on the affirmation side to the value on the negation side to identify the time-series pattern on the time-series data, in the time-series pattern identification process.

14. The non-transitory computer-readable recording medium storing the information analysis program according to claim 11 , which causes the computer to execute a process that further checks whether or not the time-series data has assumed a divergence state in which a variance value of the affirmation level of the time-series data for each time interval exceeds a first threshold value defined in advance or the time-series data has assumed a convergence state in which the variance value falls below a second threshold value defined in advance to identify the time-series pattern on the time-series data, in the time-series pattern identification process.

15. The non-transitory computer-readable recording medium storing the information analysis program according to claim 11 , which causes the computer to execute an effective period confidence measure calculation process that extracts the time-series data included in the effective period identified in the effective period identification process, and then calculates a confidence measure of the time-series data included in the extracted effective period.